A vision-based controller for path tracking of autonomous underwater vehicles

Abstract

This paper aims a design of a vision-based controller for autonomous underwater vehicles (AUVs). The control system performs the path following of a line on the seafloor and tunes automatically the cruise kinematics for the advance.The navigation process rests upon a sensor which estimates the relative vehicle position from the perspective of the camera directly without the need of employing camera models for estimating physical variables. For the state estimation, image processing in real time is employed. The control strategy is developed in the state space of pure vision coordinates. The strategy points to reach rapidly favorable configurations of the AUV and afterwards to pursue a perfect alignment with the line. Experiments of the real subaquatic world are presented.